Search papers, labs, and topics across Lattice.
5
0
8
7
Stop letting SFT ruin your LMMs: PRISM uses on-policy distillation to realign your model *before* RL, boosting performance by up to 6%.
Hallucination mitigation in LVLMs doesn't have to come at the cost of general performance: MPD reduces hallucinations by 23.4% while *improving* overall generative capabilities.
Instance-specific timestep schedules can significantly boost diffusion model performance, challenging the reliance on global discretization strategies.
Ditch the handcrafted coefficients: DyWeight learns how to dynamically weight gradients in diffusion model sampling, slashing compute while boosting image quality.
Forget retraining: ACE-Merging unlocks state-of-the-art model merging by cleverly estimating task covariance from parameter differences alone.
